JOURNAL AND PROCEEDINGS. 77 
Let, now, light enter the polariser from a powerful source (arc 
lamp). It emerges polarized. ‘Then let it pass through the analy- 
ser and be received in the eye or on a screen. We can turn the 
analyser so that no light can get through, and the screen is dark. It 
must be remembered that the analyser and polariser, when examined 
in common light, are perfectly transparent, and yet when placed one 
after the other can completely shut out the light. If the analyser is 
turned either way, the light gets through and illuminates the screen. 
When arranged so that no light gets through, polariser and analyser 
are said to be crcssed; when turned, that maximum light gets 
through, they are said to be parallel. 
Suppose, now, polariser and analyser are crossed ; the screen is 
dark. On inserting between polariser and analyser a piece of mica 
or any other thin crystalline plate, very beautiful colors are seen on 
the screen. These are due to interference. When the white light, 
polarised by the polariser, enters the crystalline plate, it is broken 
up by the crystal into two portions, each polarised, one with vibra- 
tions in a certain direction, the other with vibrations at right angles 
to these, these two directions being determined by certain lines in 
the crystal. Now, one portion travels faster than the other, and so 
when they emerge on the other side of the plate the latter lags 
behind. At once comes the suggestion that the two portions should 
interfere. But they are polarized in planes perpendicular to each 
other, and as motions at right angles can never destroy each other, 
they cannot. And so we make them pass through the analyser which 
brings them together again and makes interference possible. But 
the various components of white light—red, yellow, blue, etc.—do 
not travel in the crystal with the same speed, and so while the waves 
proper to one color interfere and cause that color to disappear, other 
colors will not go out, and we get on the screen what is obtained by 
subtracting certain components from the white light. Again, if 
the crystalline plate is not of uniform thickness, there will be seen a 
color proper to each thickness. Hence by making a pattern by 
using sheets of mica of varying thickness, we can show any figures 
we desire on the screen in gorgeous colors. 
By rotating the analyser through go* the colors become com- 
plementary to the former ones. Every sheet of doubly-refracting 
crystal shows these effects, but mica is the easiest to work with. 
